national governments when the latter’s policies deviate from the guidelines. If countries do not respond to the directives, the Council may recommend that the European Investment
Bank stop lending to the country, require it to make noninterest bearing deposits with the Community and impose fines. The Stability and Growth pact, 1995, emphasizes the
importance of maintaining sound government budgetary discipline as a means to strength- ening the conditions for price stability and strong sustainable growth conducive to employ-
ment creation. The rules of the stability pact were aimed at keeping the budget deficits within 3 of GDP. According to the “excessive deficits” criteria, sanctions and fines would be
imposed on member countries violating the 3 deficits to GDP criterion. In the context of my model, this corresponds to a situation in which the fiscal authorities act in a committed
fashion and adhere to rules regarding the tax rates imposed in their respective countries.
The temptation to use monetary policy for short-term gains in employment or reductions in real value of government debt causes credibility problems vis-a-vis the private sector. This
situation is represented by the monetary discretion scenario in the model developed in the paper. On the other hand, the central bank could be made independent of the government and
the electorate and be given a reward for maintaining price stability. The monetary commit- ment scenario considered in the paper reflects the realization of this outcome. Allan Drazen
1989 shows that free capital markets and inflation convergence would force the European countries to restructure their tax revenue base, replacing inflation taxes with more direct
taxes. My paper considers a tax effect and seigniorage effect, which reveals the likely impact on government spending.
The paper is organized as follows. The two-country framework with a common monetary authority is presented in section II. In section III.A and III.B, the solutions for inflation,
output and public expenditure deviations under symmetric and asymmetric scenarios of rules and discretion are analyzed, respectively. This reveals the interactions among fiscal distor-
tions and monetary and fiscal policy time-inconsistencies in a two-country monetary union. Section IV deals with the scenarios of rules and discretion under fiscal coordination.
Concluding observations are made in section V.
2. Model
Following VanHoose 1992 and Bryson et al. 1993, the model is a two-country extension of the closed-economy framework of Alesina and Tabellini 1987. As in Bryson
et al. 1993, the two fiscal authorities choose the respective tax rates for each country. The overall composite inflation rate is a choice variable for the monetary authority, but it can be
expressed as a weighted average of the endogenously determined inflation rates in the two countries. Their individual inflation rates adjust endogenously to the monetary authority’s
decision about the overall inflation for the union. The government budget constraint used in the model is a modified version of that used in Alesina and Tabellini 1987 and Bryson et
al. 1993, in which the government expenditure g is represented as the ratio of nominal government spending to nominal income. A log linear production function is considered,
such that output may be expressed in terms of the logarithm of equilibrium employment level.
47 G. Banerjee Journal of Economics and Business 53 2001 45– 68
The following two-country framework is as follows:
P
t c
5 fP
t
1 ~1 2 fP
t
1 where P
t c
is a composite inflation rate; and P
t
and P
t
are the endogenously determined inflation rates for the prices of goods produced in countries 1 and 2 respectively. In each
country, workers consume a fraction, b of home country output and a fraction 1 2 b of foreign country output. A fraction f of country 1 inflation and a fraction 1 2 f of country
2 inflation contributes towards the composite inflation. The common monetary authority sets the composite inflation rate for the two countries. The fraction f is chosen proportional to
the size or population of country 1 relative to country 2, so that more weight is attached to the endogenously determined inflation of the country having a larger population. The
government budget constraints faced by the countries’ fiscal authorities are:
g
t
5 t
t
1 VP
t c
2a g
t
5 t
t
1 ~1 2 VP
t c
2b where V is the fraction of common monetary seigniorage allocated by the common monetary
authority to country 1. The fraction would depend on the size of the country in the Union. These approximations to the government budget constraints follow Alesina and Tabellini
1987, by abstracting from the intertemporal dimension of the government budget constraint through the assumption that government expenditures are not financed by issuance of public
debt.
1
Government expenditures are determined once tax rates and money seigniorage have been chosen. Unlike Alesina and Tabellini, there is a common money seigniorage for the two
countries, which is determined by the overall inflation rate set by the single central bank. The common monetary seigniorage is divided between the two countries. Government expendi-
ture is financed partly from taxes and partly from the seigniorage revenues.
As in Duca 1987 and Duca and VanHoose 1990, the workers value real wages they earn in terms of the aggregate consumer price index, which accounts for the price level in
both the countries, as they consume output from both the countries. In contrast, the firms value the real wages they pay in terms of the prices of the products which they produce.
Thus, the aggregate consumer price index enters the labor supply functions while home price levels enter the labor demand functions. The labor supply functions for country 1 and 2 are
expressed as in Duca and VanHoose 1990 in country 1 and 2 respectively:
Country 1 Country 2
Output 5 y
t
Output 5 y
t
Price Level 5 p
t
Price Level 5 p
t
Ratio of government spending to nominal income 5 g
t
Ratio of government spending to nominal income 5 g
t
Tax Rate 5 t
t
Tax Rate 5 t
t
CPI: pˆ
t
5 bp
t
1 1 2 bp
t
CPI:pˆ
t
5 bp
t
1 1 2 bp
t
48 G. Banerjee Journal of Economics and Business 53 2001 45– 68
l
t s
5 c~w
t
2 pˆ
t
3a l
t s
5 c~w
t
2 pˆ
t
3b where w
t
and w
t
represent the log of nominal wage in country 1 and 2 respectively. On the labor demand side, firms choose the quantity of labor employed to maximize their profits. As
in Bryson et al. 1993, the domestic fiscal authorities tax the sales of domestic firms as well as foreign firms at the domestic tax rate,
2
within its borders. Using the first-order condition for profit maximization, the labor demand functions are:
l
t d
5 2b~w
t
2 p
t
1 bt
t
1 ~1 2 bt
t
4a l
t d
5 2b~w
t
2 p
t
1 ~1 2 bt
t
1 bt
t
4b where b 5 11 2 a . 0 and 0 , a , 1. Labor demand depends negatively on real wages
and tax rates. The coefficient ‘a’ is obtained from the production function y 5 al. The derivation of the labor demand functions is obtained in App. A of Bryson et al. 1993.
The aggregate supply functions for the two economies follow from a nominal wage contracting approach. The details of the calculation are available from the author upon
request. First, labor supply is equated with labor demand to obtain the Walrasian, full-information wage. The contract wage is equal to the expected value of the full-information wage. The contract
wage is then substituted into the labor demand function, and output is obtained by substituting the labor demand function into the log-linear production function y
t
5 al
t
: y
t
5 ~ab 1 cb
2
~P
t
2 P
t e
1 bc~P
t
2 ˆ P
t e
2 b
2
b~t
t
2 t
t e
2 bcbt
t
2 b
2
~1 2 b~t
t
2 t
t e
2 bc~1 2 bt
t
} 5b
y
t
5 ~ab 1 cb
2
~P
t
2 P
t e
1 bc~P
t
2 ˆ P
t e
2 b
2
~1 2 b~t
t
2 t
t e
2 bc~1 2 bt
t
2 b
2
b~
t
2 t
t e
2 bcbt
t
} 5b
In Eq. 5, overall CPI inflation is P ˆ 5 bP 1 1 2 bP. Taking expectations and
rearranging terms we obtain: ˆ P
e
5 1 2 bz
e
1 P
e
, where the real exchange rate depreci- ation is defined by: z 5 P 1 e 2 P, and e is the rate of nominal exchange rate depreciation
expressed in terms of home currency per unit of foreign currency. In case of a common currency, we can consider e 5 0 in logs. The domestic aggregate supply has a negative
relationship both with the domestic and the foreign tax rates. The derived labor demand for domestic firms depend on sales at home and abroad. As a result, the domestic firms
production may fall due to a rise in foreign tax rates. The nominal income equilibrium condition shows the relationship between nominal income and desired private nominal
spending as well as government spending. Following Bryson et al. 1993, it may be represented as follows:
P
t
Y
t
5 ~1 2 t
t
P
t
Y
t b
P
t 211u
~1 2 t
t
P
t
Y
t 12b
~EX
t
P
t u
1 G
t
6 where EX denotes nominal exchange rate in terms of home currency per unit of foreign
currency E 5 1, in case of common currency, G is government spending and u . 0 shows
49 G. Banerjee Journal of Economics and Business 53 2001 45– 68
the degree of market integration and increases with greater integration of markets. Dividing by P
t
Y
t
on both sides of Eq. 6, and taking logs we get after simplification the implied real exchange rate depreciation similar to Bryson et al.:
z 5 1 2 b
u P
t
2 P
t
2 t
t
1 t
t
1 y
t
2 y
t
7 The equilibrium conditions for domestic and foreign goods, which follow from above
nominal income equilibrium condition, are given by y 5
~u 1 1 2 b 1 2 b
~P 2 P 1 y 1 ~t 2 t 8a
y 5 ~u 1 1 2 b
1 2 b ~P 2 P 1 y 1 ~t 2 t
8b Substituting for the CPI inflation in Eq. 5a, we obtain the domestic aggregate supply
function in terms of the implicit real exchange rate depreciation: y
t
5 ~ab 1 c~b
2
1 bc~P
t
2 P
t e
2 b
2
b~t
t
2 t
t e
2 bcbt
t
2 b
2
~1 2 b~t
t
2 t
t e
2 bc~1 2 bt
t
2 bc~1 2 b z
e
} 9a
Similarly, the aggregate supply in country 2 may be expressed in terms of the expected real exchange rate depreciation:
y
t
5 ~a~b 1 c~b
2
1 bc~P
t
2 P
t e
2 b
2
~1 2 b~t
t
2 t
t e
2 bc~1 2 bt
t
2 b
2
b~t
t
2 t
t e
2 bcbt
t
1 bc~1 2 b z
e
} 9b
Eqs. 9a and 9b are similar to Bryson et al. They differ from Alesina and Tabellini in that both monetary and fiscal policies are subject to potential time inconsistencies. In terms
of Eq. 9, we notice that the higher the actual inflation rate is relative to the rate expected by the private sector, the higher is output and employment. Again, the lower the actual tax
rate is relative to the expected tax rate, the higher is output.
Full-information output is the output obtained in the absence of any uncertainties, or, in the context of Eq. 9, the output when inflation, tax and real exchange rate are equal to their
expected values.
y
f
1 a
~b 1 c 2 bcbt
t
2 bc~1 2 bt
t
2 bc~1 2 b z 10a
y
f
5 a
~b 1 c 2 bc~1 2 bt
t
2 bcbt
t
1 bc~ 2 b z 10b
The desired or target output level in each nation is assumed to be the level of full-information output without any tax distortions. The taxes tend to reduce the level of output below its ‘natural’
level. The nondistorted output in each country may be expressed as follows:
50 G. Banerjee Journal of Economics and Business 53 2001 45– 68
y¯ 5 2 abc
~b 1 c ~1 2 b z
11a y¯ 5
abc ~b 1 c
~1 2 b z 11b
The policy makers aim to minimize the deviation of actual output from the desired or nondistorted output. Combining Eqs. 9 and 11, we get the following expressions for
deviations of output from the nondistorted or target level in countries 1 and 2, respectively:
~ y 2 y¯ 5 ~abb 1 c~b 1 c~P
t
2 P
t e
2 bb~t
t
2 t
t e
2 cbt
t
2 b~1 2 b~t
t
2 t
t e
2 c~1 2 bt
t
2 c~1 2 b~ z
e
2 z} 12a
~ y 2 y¯ 5 ~abb 1 c~b 1 c~P
t
2 P
t e
2 b~1 2 b~t
t
2 t
t e
2 c~1 2 bt
t
2 bb~t
t
2 t
t e
2 cbt
t
1 c~1 2 b~ z
e
2 z} 12b
Similarly, the policy makers have a desired level of public expenditure, g¯, which may be higher than the actual level of government spending g. The authorities aim to minimize the
discrepancy between the actual and desired level of public expenditure, and this is incorpo- rated in their loss functions. The common monetary authority seeks to minimize the
following loss function:
V
MA
5 1
2 ~P
c 2
1 m
1
H
C 2
~ y 2 y¯
2
1 2 2 C
2 ~ y 2 y¯
2
J
1 m
2
H
C 2
~ g 2 g¯
2
1 2 2 C
2 ~ g 2 g¯
2
J
13 where, m
1
. 0, m
2
. 0, and g¯ 5 target government spending .0. Time indices have been omitted for notational convenience. The monetary authority seeks to minimize the deviation
of common inflation rate, P
t c
, from a goal of zero, departures of output from nontax distorted output, and deviations of public expenditure from the target government spending. The loss
weights on output and public spending deviations from target are assumed to depend on the relative size of the countries in terms of their population or output.
The fiscal authorities choose the tax rates in their respective countries to minimize the following loss functions:
V
FA
5 1
2 ~P
c 2
1 d
1
~ y 2 y¯
2
1 d
2
~ g 2 g¯
2
14a V
FA
5 1
2 ~P
c 2
1 d
1
~ y 2 y¯
2
1 d
2
~ g 2 g¯
2
14b As in Alesina and Tabellini, the fiscal authorities are assumed to attach at least as much
weight on the output and government spending objective relative to inflation as compared to
51 G. Banerjee Journal of Economics and Business 53 2001 45– 68
the monetary authority i.e. d
1
,d
1
m
1
and d
2
,d
2
m
2
. Therefore, the fiscal authorities’ concerns about output and public expenditure deviations from target levels are at least as
great as those of the common monetary authority. The member countries in the EMU differ in respect to their economic structure, macroeconomic performance, and fiscal policy
preferences. The effects on national macroeconomic performance due to asymmetries in the national fiscal preferences, d
1
Þ d
1
; d
2
Þ d
2
, and the desired levels of government spending, g Þ g have been analyzed in the context of a two-country monetary union by B. Van Aarle
and Huart 1999.
3. Policy making under rules and discretion
